TCP for Home Multimedia:Why you can’t teach an old dog new tricks

Hariharan RahulSzymon Chachulski, Kah Keng Tay,

Dina Katabi

TCP’s shining moment

• NSFnet backbone crumbled in 1986• Throughput dropped from 32 kbps to 40 bps

Congestion Control saved the day!

A thousand cuts since!

• High-delay networks – Satellite links

• High-bandwidth links– 100 Gbps to Terabit fibers

• Real-time traffic– Gaming, IP TV

• Datacenters and cloud computing

TCP under siege

You are an analog computer in a digital world!

The last straw: Wireless Multimedia

• Home Entertainment to grow to $12B by 2010 – Jupiter Research

• Multimedia home networks growing at 46% compounded – Frost and Sullivan

Why should TCP change?

It’s not what TCP does; It’s how it does it!

• Wireless is lossy Needs loss recovery• Wireless is a scarce shared resource Needs

congestion control

But…

TCP’s Architecture Is Too Rigid

• Ignores the characteristics of the higher layer– Provides complete reliability regardless of what

the application needs• Ignores the characteristics of the lower layer– Congestion control reacts to all losses, regardless

of their cause

Live Video StreamingTCP Video Packets

TCP ACKs

Server Client

• Server and client using 802.11b• VLC for video streaming over TCP• Asymmetric links

– Forward link good– Reverse link poor

1234s

Live Video StreamingTCP Video Packets

TCP ACKs

Server Client

7.00 9.00 11.00 13.00 15.00 17.00 19.00 21.00 23.000

50

100

150

200

250

300

350

Time (secs)

Thro

ughp

ut (K

B/se

c)

Burst ACK Loss

TCP recovered,Video still frozen

Video recovers

TCP ignores higher layer needs and lower layer characteristics!

TCP’s Architecture Is Too Rigid

• Ignores the needs of the higher layer• Ignores the characteristics of the lower layer• Couples logical and physical packets

Multicasting Video

• Many popular applications– Mobile TV– Security videos in airports and train stations– Commercials or music videos in malls and

nightclubs• But wireless multicast needs loss recovery!

The Multicast Experiment

• Lecture streamed via an access point• All nodes use 802.11b• Nodes simultaneously subscribe to lecture

video

Many Unicasts Congest the Medium

Capacity

User 1 ReTxUsr1 User 2 ReTx

Usr2 User 3 ReTxUsr3

Wastes the fundamental broadcast advantage of wireless!

Smarter Multicast Scales Better!

Capacity

Common ReTxUsr1

ReTxUsr2

ReTxUsr3

ReTxUsr4

ReTxUsr5

ReTxUsr6

TCP’s coupling of physical and logical packets hinders wireless

multicast!

TCP’s Architecture Is Too Rigid

• Ignores the needs of the higher layer• Ignores the characteristics of the lower layer• Couples Logical Connections with Physical

Packets• Intertwines reliability and congestion control

TCP Stifles Innovation

• Mesh networks for a multistory home• Innovative opportunistic protocols for high

throughput • But do not work with TCP!

ChoiceTCPLow

Throughput

InnovationHigh

Throughput

Mesh Network Experiment

• 2-hop mesh network running 802.11b• Determined throughput sustainable by– TCP– Opportunistic Routing Solutions

• Encoded video at the corresponding rates• Transmitted encoded video over mesh

network

TCP’s intertwined reliability and congestion control stifles innovation

Conclusion

• We need TCP’s functions!• But TCP’s architecture shackles us!– Rigid layering does not understand application

needs or medium behavior– Tight coupling of physical and logical packets not

conducive to multicast– Intertwined reliability and congestion control stifle

innovations for high throughput

The time has come for a newer, nimbler alternative!